医学图像分割及其发展现状

State-of-the-Art in Medical Image Segmentation

医学图像分割是各种医学图像应用的基础,当前的临床辅助诊断、图像引导的外科手术和放射治疗中,医学图像分割技术显示出越来越重要的临床价值.由于医学图像种类繁多,常规影像包括磁共振（MR）成像、计算机断层（CT）成像、正电子发射计算机断层显像（PET）、超声（US）成像等,其中MR成像还可以产生多种不同时间参数序列的图像模态.为此,医学图像分割技术已成为面向不同的影像模态、临床目标、特定解剖学部位的一种独特的应用科学体系.结合现有的国内外研究成果,该文详细地介绍和系统地对比了图像分割方法并进行了分类,最后还对6个国际知名医学成像期刊和会议进行了统计分析,阐述了医学图像分割技术的研究趋势.

Medical image segmentation is the basis of various applications of medical image,which have shown more and more important values in the current clinical computer-assistant diagnosis,image guided surgery and radiological treatment.Many categories of medical images have been researched for clinical requirement such as the imaging systems of magnetic resonance（MR）,computer tomography（CT）,positron emission computed tomography（PET）and ultrasound（US）,where MR imaging can produce image modalities with different time-parameter series.Therefore,technologies of medical image segmentation became a particular applied science system which serves for different imaging modalities,clinical object and anatomic site.Integrating with the existing research results from home and abroad,the paper gave detailed introduces and contrasts for current methods of medical image segmentation;the development trend of medical image segmentation technologies was also shown,based on investigation and analysis of six most prestigious scientific journals and conferences.